We have both old and more recent data which indicate that adults at high-risk for HIV infection sometimes become transiently infected with HIV. This includes: (1) transient appearance of serum antibodies (Ab) to conformational epitopes to gp41/gp160; (2) transient appearance of HIV in the PBMCs at the same time as Ab in many cases; (3) paternity typing of PBMC samples taken at the time of Ab and virus positivity and later when the subject is negative for both, showing that they are from the same subject; (4) sequence analysis of the viruses present in transiently infected subjects, showing that they are not lab contaminants; and (5) demonstration in one case in which the virus in a transiently infected subject was genetically related to the virus in genital fluids from his HIV+partner. Longitudinal statistics from studies of 37 high-risk subjects over 4 years showed that transient infections were more frequent than progressive infections, and that the former was associated with unprotected oral sex, while the latter was associated with unprotected genital sex. HIV was found by PCR in the tonsil of one such transiently HIV infected subject, but it was not detected in the PBMCs. We hypothesize that transient HIV infections do occur in high-risk adults and that this is due principally to the route of infection, i.e. oral infection. However, we will focus this project only on proving the hypothesis that transient infection occurs. Our specific Aims are: (1) to obtain overwhelmingly convincing documentation of cases of transient HIV infection; and (2)determine if monitoring for HIV in tonsil biopsies and anti-gp41 Ab IgA, IgG and IgM in saliva is better than or provides more information than monitoring in blood. Although transient infection is presently controversial, our data strongly indicate that it occurs and that it may be due to oral infection. [unreadable] [unreadable]